Marketing personalization is one of the most important elements for marketers to concentrate on in the next few years. In a study from Adobe titled 'Digital Roadblock: Marketers Struggle to Reinvent Themselves’, one third of participants named it as the most important development in the future. They are already having some impact and we have seen companies like Amazon reap huge benefits from adopting the approach through suggestion engines and personalized emails.
It is unsurprising that retailers are at the forefront of this revolution, but according to research from Gartner, only 10% of tier 1 retailers classified themselves as 'highly effective' at it.
This lack of confidence in marketing personalization comes predominantly from companies not using data to its full effect. Khalid Khan, head of analytics at A.T. Kearney believes that 'Today it's about one size fits many. Organizations overlook very important behavioral differences and differences in need states - what people might want at that particular point in time. The ability to collect and mine the right data is the biggest challenge now.' The problem that most companies have is that they are unable or unwilling to do this.
In most companies' current approaches, they make several assumptions about people which essentially silos them into relatively large groups. It could be that all who say they are are a ’father with an 11 year old son' are grouped into a single silo, but this approach is only faux-personalizing. There may be two fathers who both have an 11 year old son, one of whom is a huge fan of football and another who hates it, but they will be receiving the same message because they are technically in the same silo.
Instead, data needs to be collected that can allow companies to truly personalize their messages to an almost macro level.
This comes down to effective use of data and identifying what you can be almost totally certain about an individual. For instance, if you know their address and the climate for the area you can send content specific to dealing with a temperature range like 'protecting your pipes from frost damage' or 'how to stop the sun bleaching your walls'.
One problem that some companies have is identifying people only on a small range of purchases that may not reflect them, so they may have bought a guitar, does that mean they like guitars or that they like buying gifts for somebody who does? Looking back through purchase history can help to clarify assumptions, and the richer the history the more specific you can be. So, if this person has bought several guitars specific to jazz, you could safely assume they like playing jazz music and then appropriately tailor content or products they will receive. If they had bought one jazz guitar then it may just be anomalous.
Essentially, the challenge that marketers are currently having with personalized data isn't that they don't have enough data, it is instead that they either don't know how to use it or they can't get it into workable enough order to utilize it effectively.
It is why the use of data scientists and big data technologies is becoming essential, as having the data is not enough for personalization, it needs to be sorted, identified and utilized in the correct way.